The 4th-generation trade-off

The market’s structural illusion.

The industry posits modern core selection as a choice between simplicity and flexibility. That framing is convenient, but is overly reductive and, as such, also misleading.

At one extreme are closed-box cores promising out-of-the-box functionality and rapid deployment: standardised products, vendor-managed upgrades and predictable release cycles. Complexity appears contained.

At the other extreme are open-ended cores that offer extensive freedom to build and customise. Differentiation appears unconstrained.

But this is a false dichotomy that hides the structural trade-offs within each approach.

The closed-box core

In the case of the closed-box cores, simplicity at the outset can become dependency over time. Product innovation becomes conditional on prioritisation cycles outside an institution’s control. If a required capability is not on the roadmap, an institution must either wait, which frustrates its ability to move quickly, or build it itself, which introduces high total cost of ownership (TCO) and complexity and undermines the strategic and economic case for a closed-box core.

Over time, strategy aligns with what the platform enables, rather than what the market demands. The platform remains structurally coherent and upgradable. But the business absorbs the constraint.

The open-ended core

With the open-ended core, the autonomy gained also comes with a trade-off. It comes with open-ended responsibility.

The ability to make deep structural customisation allows an institution a massive level of autonomy to introduce new logic and adapt the system to its own idiosyncratic use cases. However, over time, the institution is no longer consuming a product; it is maintaining its own variant of it.

Upgrade cycles slow down. Version adoption depends on regression effort. In extreme cases, upgrades are deferred altogether. Innovation continues locally while structural divergence grows.

Flexibility quietly becomes technical debt.

The middle ground

We believe that the right 4th-generation core offers a middle-ground between these two extremes that solves for the corresponding trade-offs.

It does not remove complexity. It governs it deliberately through configuration and controlled extensibility, which allow scope for change, data enrichment, and new functionality, but within defined boundaries.

As a result, upgrades remain operationally viable rather than theoretically possible. Institutions can iterate products and absorb regulatory change without accumulating structural debt or latency. This discipline becomes even more critical in an environment shaped by artificial intelligence (AI) and increasingly by agentic workflows that act autonomously across systems.

AI will increasingly influence customer interaction: predictive servicing, dynamic pricing, proactive risk intervention, real-time credit decisioning. These capabilities do not sit inside the core. But they depend on it.

AI requires clean, real-time data flows and clearly exposed domain events. It depends on the separation between transaction authority and surrounding intelligence layers. It requires models to evolve continuously without destabilising ledger integrity. And it relies on the context provided by rich metadata.

Therefore, the right 4th-generation core is one that balances simplicity and flexibility by enabling high autonomy without accumulating structural debt. But the right 4th-generation core is also one that serves as the foundation for assembling the right ecosystem of third-party capabilities, offering real-time data and events that enable the institution to scale its operations, adapt at pace, and capitalise on the promise of AI to deliver much more personalised customer service.